Harm, I would usually use a DATA step and a retain statement for that.  Something like DATA want; SELECT have; BY category; RETAIN counter;  DROP counter; IF first.category THEN counter=0; IF counter <=3 THEN OUTPUT; RUN; Doc ... View more

Check the log for your code.  EG typically puts some wrappers on the code you write and it may be causing the effect you see.  The ODS graphics part of the wrapper would be based on the graphics settings you have in tools --> options.  Doc Muhlbaier Duke ... View more

One approach would be: Assign a pseudo-random number between 0 and 1 (the RANUNI function) to every record. Sort by customerid and that number (the category will now be in a random order within customerid). Select the first of each customerid (now have one record per customerid). Re-sort by category and the random number (the customerids will now be in a random order within category). Select first 3 in each category. Doc Muhlbaier Duke ... View more

The answer from Anotherdream and your last statement, "or for the length specified by informat", are consistent.  Type has a "format" and Quest1 has an informat.  They are different in use, but have the same tools. Doc Muhlbaier Duke ... View more

"No merit" is a bit strong a term.  Once you get estimates, you still have to decide how to use them and the Bayesian and classical frequentist interpretations can still be different. ... View more

You can output the coefficients using the ODS capabilities.  See the ODS Output section of the SAS documentation for your version of PROC GLM.  In 12.1, the table is named ParameterEstimates. ... View more

You will need to tell us more... Are you using an Enterprise Guide task or writing code?  What is the error message?  Can you include the log?  What is the version number and platform for your SAS server? ... View more

With large N, the maximum likelihood estimates for a Bayesian and frequentist analysis should converge to the same estimates, as the data will overwhelm any prior.  (In logistic regression analysis, the effective sample size is determined by the group whose % is nearest the boundary so it could be that your effective sample size is much smaller than 30,000 and my comment won't help.). Doc Muhlbaier Duke ... View more

LM, I see that this is your first post.   You will likely get answers quicker if you post to the forum that is nearest to the topic of your question rather than this catch-all one.  "SAS GRAPH and ODS Graphics" would be ideal for this question. Validation means different things in different contexts.  When you say 'validate' do you mean that the graph looks like it is supposed to?  Or are you asking a more general question about the performance of the SAS system?  Or something else?  For general validation, I use the "SAS Installation Qualification" tool to make sure that all of SAS is working as intended. If you looking for whether a specific picture is correct, my recommendation is that you search the support.sas.com web site for an example of what you want and run that data to see if you get the same results. Doc Muhlbaier Duke ... View more